Re: [boost] Re: no semaphores in boost::thread

From: William E. Kempf (wekempf_at_[hidden])
Date: 2003-07-07 09:25:35

Jon Biggar said:

> There is actually one case that needs a semaphore that has no reasonable
>  alternative in pthreads.  The only pthread synchronization operation
> that is asynch-reentrant safe (i.e. can be called from a signal handler)
>  is signaling a pthread semaphore.
>
> It would be nice if there was some abstraction available in boost
> threads that used this mechanism internally to get the needed behavior,
> but encapsulated to hide the error-proneness of a raw semaphore.

The problem is that both your problem case and your solution are POSIX
specific.  I can't envision an interface that would be both portable and
usable on other platforms.  Provide one and I'll certainly consider it.

-- 
William E. Kempf